In [127]:
%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt

We need import matplotlib for plotting and numpy for array operations.

In [128]:
from skimage import data
import imageio

photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')

And also we need import data from skimage and imageio for read and manipulate pictures

In [129]:
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[129]:
<matplotlib.image.AxesImage at 0x7fc20cf3b4e0>

We use ply.figure() for resize image and plyt.imshow() for showing image on the screen

In [130]:
photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')
print("Shape of photo_data:", photo_data.shape)
low_value_filter = photo_data < 100
print("Shape of low_value_filter:", low_value_filter.shape)

Shape of photo_data: (3725, 4797, 3)
Shape of low_value_filter: (3725, 4797, 3)

Import the picture again and make a low_value_filter which pixel values (rgb) less than 100

In [131]:
photo_data[low_value_filter] = 70
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[131]:
<matplotlib.image.AxesImage at 0x7fc20ce947b8>

Using low_value_filter I reassign all the pixel values which are less than 100 to 70.

And we have a brighter image on the screen

In [132]:
photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')
blue_mask  = photo_data[:, : ,2] < 100

photo_data[blue_mask] = 0
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[132]:
<matplotlib.image.AxesImage at 0x7fc20cc19cc0>

For now I use blue_mask which is photo_data[:, : ,2] < 10

Here, I pick all the row and columns which are blue pixels (0:R, 1:G, 2:B) and less than 100 with using index slicing and I reassign all the blue pixel values which are less than 100 to 0. 0 means all the pixels are being black.

In [133]:
photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')
red_mask   = photo_data[:, :,0] < 100

photo_data[red_mask] = 0
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[133]:
<matplotlib.image.AxesImage at 0x7fc20cb78e10>

For now I use red_mask which is photo_data[:, : ,0] < 100

Here, I pick all the row and columns which are red pixels and less than 100 with using index slicing and I reassign all the red pixel values which are less than 100 to 0. 0 means all the pixels are being black.

So we have a darker image which devoid of red.

In [134]:
photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')
green_mask = photo_data[:, : ,1] < 75

photo_data[green_mask] = 0
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[134]:
<matplotlib.image.AxesImage at 0x7fc20cb53f28>

For now I use green_mask which is photo_data[:, : ,1] < 75

Here, I pick all the row and columns which are green pixels and less than 75 with using index slicing and I reassign all the green pixel values which are less than 75 to 0. 0 means all the pixels are being black.

In [135]:
photo_data = imageio.imread('/home/mert/Desktop/Data_Science/Week-3-Numpy/filter.jpg')
red_mask   = photo_data[:, : ,0] > 100
green_mask = photo_data[:, : ,1] > 75
blue_mask  = photo_data[:, : ,2] < 90

final_mask = np.logical_and(red_mask, green_mask, blue_mask)
photo_data[final_mask] = 0
plt.figure(figsize=(15,15))
plt.imshow(photo_data)
Out[135]:
<matplotlib.image.AxesImage at 0x7fc20cab7780>

Here, I merged all the masking features with using np.logical_and function. And we have a mixed masking picture on the screen.